A bigger budget, up to a point, will give a more accurate result, but this can be partly overcome by doubling up for a week to let the system build some benchmarks and then cutting back to level out your overall spend.
You definitely don’t need hundreds of conversions to learn about the potential for lead generation. If on the other hand you want an end to end assessment of either ecommerce from click to cart to paid customer or lead to sales process to paid customer, then you need more planning, possibly more people (sales) and more budget. I would caution here that unless you are very confident about your salesforce, it is often wise to test end to end, especially if they are used to working different type of leads.
When calculating budget for a pilot, we usually give you a recommended range of budgets based on the available impressions/traffic and cost per click across a range of keywords.
Two things about a trial that are important are that a. Budgets are always a rule of thumb estimate, precision simply can’t be done and will waste a lot of effort. That doesn’t mean we can’t set a ceiling and guarantee not to go beyond it, but that we can’t be certain it is enough to answer your question, just that it is our best guess.
How many conversions would give a good trial?
This depends on your market niche and how you define a conversion.
Ideally you need to consider each product an individual trial otherwise there are too many variables that might come in to play and fool you.
If you have one conversion goal – like a subscription service or lead then you will need fewer conversions to learn how well PPC could work it is unlikely to be less than 20 and to understand the quality of leads you probably need 50 or 60 minimum.
For an ecommerce website you need a good number of conversions per product or at least product group. It is different for every niche and depending on your question.
As this is a rough calculation, don’t get too concerned with getting this figure right. Just get a feel for how many conversions you’ll need based on the numbers above.
Research keywords
Because this is a rule of thumb guess, you don’t need a huge list of keywords,
Google provides a free keyword planner online and there are any if you search.
For a trial its important to keep it simple. Keywords should be unambiguous as possible. Avoid any that might be misleading, e.g Match could football, barbecue, or a dating site. Leave this complexity for another time.
- Check the projected spend on adwords planner to make sure your word will get searched, otherwise it is pointless.
- Searching in Google for your keyword tells you if others are advertising. Competition is a good sign.
Cost per click estimates will give an idea from which to project conversion costs.
Estimate your conversion rate
This chart form unbounce gives useful industry averages that you can fall back on for the purpose of initial calculations.
If your industry is not there, just grab the one most similar, or even take two add them together and divide by two. Being uber accurate is not the goal here, because too much depends on how your landing pages convert.
Good landing pages can convert above 10% and frequently do after a little tweaking whereas the figures in the chart represent visits to random web pages that wont be very well optimised to convert.
For a trial its only about learning what you can achieve and then calculating what can be improved and what the potential scale is, so don’t get too hung up at this point.
Estimating the PPC budget for your pilot campaign
Here is how you work out the budget requirement
Conversions needed / Click Through Rate * CPC
So previously we estimated how many conversions we need let’s say 35 for this example.
Then we found that the industry average CTR (click through rate) is 3.5%
Google estimated the CPC from £8 -£12
For 35 conversions given a 3.5% Conversion rate we need 100 clicks at average £10
In this case we need £1000 to drive 35 conversions. If these are leads the cost per lead is 100/35= £28.60
Now you just need to test lead conversion and you have the full cost model.
This model will always be very much below that which you can achieve after 8 to 12 months of refining every part of the process and small improvements can have hue impacts on profitability. We always reduce CPC significantly via targeting, we always improve Conversions significantly y designing and testing the best landing pages and we can help you improve your conversions significantly if you need help with that.
Net cost of a trial
What we’ve calculated here is the advertising spend recommended to run a pilot but net cost could be a much lower and sometimes it shows a small profit immediately.
After a single month, we can usually build a model showing modes improvements over time to show it go quickly to breakeven and into solid profit.
Our recommendation is that a minimum trial be over two months followed by a full review. If it seems to be going exceptionally well, we still caution against jumping into the unknown, but to build steadily based on good data.
The final unknown in your trial and a very important one is how your competitors will react when you begin hogging the searchers and generating good business. Just like we do ever day for our clients, their agency or inhouse person will probably analyse your keywords and adverts and a whole lot more and come back with a stronger campaign to grab back a share of the market. It all depends on who your competitors are and how digital savvy they are.
In Summary
This is a method for pilot campaigns at beginners or small firms whereby they don’t already have a big brand and is only suited to eCommerce if you have no more than a hundred or two products. If you have any doubts or would simply like some help, feel free to all us on 0844 8842310 or email us on pilot@nv8v.com